EOF

Maybe the biggest thing that ever happened to Linux—at least scale-wise—is virtualization. As I recall, virtualization first materialized in a big commercial way with IBM, which started by putting many Linux instances on System z mainframes. (Once on the old Linux Show, we had a guest geek from IBM who said it was not only his idea, but also that he came up with it over lunch.) IBM didn't call those mainframes “clouds”, but that's what it hosted. Now we have clouds of clouds of Linux all over the place. Nothing could be more widespread and ordinary. (Of Netcraft's ten most reliable hosting company sites for June of this year, eight are Linux and two are FreeBSD: news.netcraft.com/archives/2013/07/01/most-reliable-hosting-company-sites-in-june-2013-2.html.)

Now think about the Internet of Things, often abbreviated IoT. It is generally assumed today that the Internet of Things will require embedded smarts. But in fact, any thing can have a cloud, whether the thing has embedded smarts or not. This insight comes to us from Phil Windley (www.windley.com), the hacker-in-chief of Kynetx (kynetx.com), a small Utah start-up with very big plans. (Disclosure: I sometimes consult them, as I do a number of other companies.) Phil is also the inventor and alpha maintainer of CloudOS, a small and simple cloud operating system for anybody and anything, including you and every thing you own. CloudOS is open source and GPL'd. So is KRL (kinetic rule language, en.wikipedia.org/wiki/Kinetic_Rule_Language), the first language for programming on CloudOS (among other things), also first authored by Phil.

By abstracting intelligence away from physical things, we can unburden those things of the need to be intelligent in themselves. In fact, we can enlarge to absolute the variety of things that can have intelligence. Phil embodies this range in the word “pico”, for persistent compute object. One of his is a pothole in front of his house (www.windley.com/archives/2013/04/potholes_and_picos.shtml). That pothole has brains in the cloud Phil gave to it, and that cloud is in Phil's personal cloud (personal-clouds.org/wiki/Main_Page). To help demonstrate how this can work in everyday life, here is a list of things I've made smart by giving each its own cloud:

Figure 1. Doc's 30" x 30" tablecloth looks like this QR code.

The doors to those things' clouds are the QR codes I've hung or stuck on them—and, in the case of the last one, the item itself is a QR code. There can be other doors as well (for example, URLs), but QR codes are handy because: a) as with Ethernet, the patent owners (Denso, in Japan) decided they would rather create more value than they capture, so they let anybody do anything they want with QR codes; and b) they're easy to scan with a smartphone.

Until now, QR codes have had the misfortune of being exploited mostly by marketers, which is why they appear as “robot barf” on promotional jive all over the place. But now it's time for the hackers to take over, which is what the ones working for Kynetx have done. They created a service (www.windley.com/archives/2013/01/introducing_squaretag.shtml) called SquareTag (https://squaretag.com), which hosts things' clouds. SquareTag's business model, for now, is selling hang-tags and stickers with QR codes on them. I'm using some of the tags and stickers for the things in the list above. (Note that SquareTag isn't a silo. I can take my things' clouds, plus my own personal cloud, and put them wherever I want. That means I can self-host them, put them in Dropbox, or stick them in some other cloud service.)

Through a feature called “safe and mine” you can present a message to any Samaritan who scans the QR code of a thing you've lost. For example, “This bag belongs to (your name). Text me at (your number).” From any computing device, you can write or change that message.

But here's the biggest thing: Every thing's cloud is a platform for relationship—between you (as the owner) and whoever else you welcome aboard: notably the companies that make and sell the things you've bought.

For example, I have messages waiting for the makers of many of the items above. Here are a few:

I also can advertise to the world, should I wish, some or all of what I say about those products in my cloud. Or, I can restrict what I say just to the companies I invite into a relationship, such as the ones above.

Likewise, any company (such as Canon, LaCrosse, Sangean and Garmin) can give every product it sells a unique cloud of its own, with its own QR code, and transfer ownership of that cloud to the customer along with the product itself. If the customer welcomes a relationship with the company, and the company agrees to the customer's terms of engagement (such as, “respect the privacy of this communication channel in the following ways”), the whole “own cycle” of a product becomes a much richer experience for both the customer and the company. The QR code then becomes what's called a “TalkTag”—meaning that its purpose is to serve as a way for the customer to signal his or her interest in talking to the company. For example, I can program the cloud of my Dish Network set-top box to make a scan of its QR code send a message to the company saying I'd like a call from an agent to help me work through a problem. I've talked to call-center people about this possibility and they love it.

What they love especially is that it's now possible to have a standard way for customers to relate to companies. The problem today is that every company's CRM (customer relationship management) system is a silo, each with its own silo'd “relationships” with customers, all of which are governed at call centers by IVR (Interactive Voice Response) systems, followed by scripted interactions when a customer gets through to a human being. If ways can be found to normalize the protocols of genuine relationships between companies and customers, the work of the call center can be made far more easy and efficient. In fact, the only way that can happen is if the customer's side does the normalizing.

Once that happens, both sides can learn far more from each other, in far better ways. If the product is the platform for a genuine two-way relationship, both company and customer are in far better positions to learn from each other. Companies can update manuals and provide notices of firmware updates. Customers can tell companies directly what's working or not working, how the product might be improved and what new products the company might consider making.

We also can start to evolve past the current marketing system, by which makers and sellers labor constantly to entrap and coerce customers into restrictive dependencies. Communications inside silo'd coercive systems tend to be far more restrictive, and far less useful, than communications between free and independent parties who are ready and able truly to help each other.

At this point, as T.Rob says in his article in this issue, we're still at the beginning of whatever it is we'll make of personal clouds. CloudOS itself is about where Linux was in the months after Linus wrote “I'm doing a (free) operating system...”. But the population of available hackers for personal cloud projects is several orders of magnitude larger than the one to which Linus wrote. Yet, the challenge is remarkably similar: take something that has been purely corporate (in this case, “the cloud”), make it personal, and then ramp it out to everything it might conceivably work on, improving it along the way, and not stopping.

The result, if all goes according to plan, is a true Internet of Things and the reframing of business around fully useful relationships between customers and companies—or, for that matter, between anybody and anything.